International Journal of ChemTech...
Transcript of International Journal of ChemTech...
Microbial production of hydrogen from sorghum stalk
T.R.Manikkandan*
*Bio Process Research Laboratory, Department of Chemical Engineering, Annamalai University, Annamalainagar – 608 002, India.
Abstract : An aerobic strain of α-ProteobacteriaAUChE 103, isolated from sorghum stalk
storage yard has been identified as a potential hydrogen producer. In the present study, the media components and process parameters were optimized for enhanced hydrogen production.
The significant media components namely glucose, malt extract, yeast extract, peptone
andNaCl were determined using Plackett-Burman design. These significant variables were then optimized using central composite design. The optimum conditions were found to be : glucose,
19.25g/L; yeast extract, 3.046g/L; malt extract, 1.64g/L; peptone, 5.640 and NaCl, 4.312g/L.
Box-Behnken design was employed to optimize the process parameters. Under the optimum
conditions a maximum hydrogen yield of 0.91 mol H2/mol glucose was achieved.
1. Introduction
The worldwide energy demand has increased due to rapid growth in population and industrial
development. To meet out the energy demand has become difficult due to fast depletion of fossil fuel reserves. Also combustion of these fossil fuel increases the green house gas emission resulting in global warming and
pollution. Hydrogen is now considered one of the alternatives to fossil fuels [1]. It is preferred to biogas or
methane because hydrogen is not chemically bound to carbon and therefore, combustion does not contribute to
green house gases or acid rainproducing only water [2]. Besides, hydrogen has a high energy yield of 122 KJ/g which is 2.75 times greater than the hydrocarbon fuel [3]. It has been reported that 50 million tons of hydrogen
are traded annually worldwide with a growth rate of nearly 10% per year [4]. Currently, 90% of commercially
usable hydrogen is obtained by steam reformation of natural gas apart from coal gasification and water electrolysis. But these processes are expensiveenergy intensive, require high operating temperatures and are
detrimental to the environment [5, 6].The other methods of hydrogen production are photocatalytic and
biological routes. Hence production of hydrogen by exploiting alternative renewable source seems to gain more
prominence. Biomass and water can be used as renewable resources for hydrogen gas production where biomass is one of the energy sources [7]. Utilization of biomass for hydrogen production through biological
means will be a dual solution for renewable source and less carbon emission process [8]. In the present study,
hydrogen production has been attempted from the renewable lignocellulosic biomass, sorghum stalk using a newly isolated strain of α-ProteobacteriaAUChE 103.
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209.
DOI= http://dx.doi.org/10.20902/IJCTR.2019.120424
International Journal of ChemTech Research CODEN (USA): IJCRGG, ISSN: 0974-4290, ISSN(Online):2455-9555
Vol.12 No.04, pp 194-209, 2019
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 195
2. Materials and methods
2.1 Isolation of hydrogen producing strain
The strain used in this study was isolated from the soil samples of maize storage yard. Pure culture was
obtained by serial dilution and plating on nutrient agar medium. The following composition: Malt 0.1% (w/v);
yeast extract, 0.2% (w/v); peptone, 0.5% (w/v); NaCl, 0.5% (w/v); glucose, 1.5% (w/v) and agar, 1.5% (w/v) at
pH 7 and temperature of 34oC was used for the growth and agar slants with 1.5% (w/v) of agar was used for the
maintenance of organism.
2.2 Acid hydrolysis
Hydrolysis was carried out by treating 5g of the powdered sorghum stalk with 100ml of 1% H2SO4 for
75 minutes at 121°C and 15psi in an autoclave. After hydrolysis, the hydrolysate was filtered through ordinary filter paper followed by filtration through Whattman No.1 filter paper. The filtrate [63.75 % (v/v) (equivalent to 1.5%
(w/v) glucose)] was mixed thoroughly with the above mentioned media composition and was neutralized with
concentrated NaOH solution to attain neutral pH.
2.3 Batch studies
The pH of the medium was adjusted to 7.0 and was sterilized in an autoclave (121oC and 14 psi) for 20
minutes. The medium was cooled and inoculated with one day pre grown culture [5% (v/v)]. The fermentation was
allowed to take place in a fermentation jar, which is kept in a constant temperature water bath in order to maintain the
fermentation temperatures. The released gas during the fermentation was collected in a separate jar by water displacement method. The gas sample was taken in a syringe and was loaded in to gas chromatography for the
qualitative assay of the hydrogen. All the experimental runs were carried out in triplicate and the average value was
taken.
2.4 Analytical methods
The gas produced during the fermentation was collected in a graduated aspirator bottle by water
displacement method at regular time intervals. The percentage of hydrogen constituted in the total gas was
determined using a gas chromatograph (AIMIL- NUCON 5765, Mumbai, India) equipped with a thermal
conductivity detector and 2.0 m (1/4 in. inside diameter) steel column filled with Porapak Q (50/80 mesh) using nitrogen as carrier gas at a flow rate of 20 ml/min. Injector, oven and column temperature was set at 150°C,
80°C and 200°C respectively.
2.5 Media optimization
The variables which significantly affect the hydrogen yield were screened using Plackett-Burman design. Fifteen variables were screened in 20 experimental runs and the insignificant ones were eliminated. The
statistical software package Minitab version 15.0 was used to analyze the experimental data. Once the critical
factors were identified through screening, the central composite design (CCD) was used to study the effect of
screened variables and their interactive effects on hydrogen production. The effect of significant medium components such as glucose, yeast extract, peptone, malt extract, beef extract and sodium chloride were tested
for their significance on the production of hydrogen.
3. Result and discussion
3.1 Isolation and identification of hydrogen producing strain
Four morphologically different colonies were picked from the agar plates inoculated with serially
diluted soil samples. They were further analyzed for hydrogen production and the strain with the highest
production was chosen for further studies. The nucleotide sequence obtained by the sequencing of PCR amplified 16S rDNA gene was compared with the available database in Genebank. Sequence analysis of 16S
rDNA gene demonstrates that the bacterial isolate is a new clone of α-Proteobacteria and designated as α-
ProteobacteriaAUChE 103.
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 196
3.2 Acid hydrolysis
Acid hydrolysis reduces the complexity of the complex substrate and the high yields of sugar after
treatment is necessary for hydrogen fermentability [9, 10]. The batch hydrolysis was carried out for sorghum
stalk in order to obtain an optimum yield of glucose. Sulphuric acid and hydrochloric acid of different concentrations ranges between 0.5% and 5% (v/v) were used in order to find an optimum acid
concentration. All hydrolysis studies were carried out in an autoclave at 121oC and 15psi. Sulphuric acid has
resulted the highest glucose yield (0.34 g glucose/g of substrate) where as the hydrochloric acid resulted a slightly decreased yield (0.30 g glucose/g of substrate). For the case of sulphuric acid the yield was increased steeply when the
acid concentration was increased from 0.5% (v/v) to 1.0% (v/v). But for the case of hydrochloric acid, the yield was
increased continuously up to an acid concentration of 4% (v/v). In both cases the yield was decreased drastically to the
minimum at higher levels of acid concentration. The decrease in yield may be due to the decomposition of glucose at higher level of acid concentrations. A higher acid concentration was also reported to increase the formation of
acetic acid and furfural, which are known for microbial growth inhibition [11]. Among the two acids sulphuric
acid with 1% concentration could be the best choice for the hydrolysis of sorghum stalk and was used throughout the experimentation.
In order to get an optimum time required for the hydrolysis of sugarcane bagasse, the hydrolysis was carried out with 1% (v/v) sulphuric acid with different hydrolysis times varied in the range of 15 to 120
minutes. The maximum glucose yield of 0.39 g/g of sugarcane bagasse was obtained at seventy five minutes
hydrolysis time after which there observed a constant decrease in the yield. The decrease in glucose yield at
higher exposure times may be due to the decomposition of the converted glucose at prolonged hydrolysis. The optimum hydrolysis conditions of 1.0% (v/v) sulphuric acid and seventy five minutes of hydrolysis time were
used for all further experimentation.
3.3 Media optimization
The medium constituents namely, carbon, nitrogen and mineral sources used in the fermentation were screened using a two-level fractional factorial experiment. Fifteen media components were investigated for their
influencein the yield of hydrogen and their coded values and the corresponding actual values are given in
Table3.1. The design was subjected to factorial analysis. Experimental and the predicted values of hydrogen are
presented in Table 3.2.Parity plot between the experimental and the predicted values showed the effects of the variables and their significance on the hydrogen yield (Fig. 1). The P values of the variables are shown in Table
3.3 and those with P values < 0.05 are considered to be significant. The Pareto chart showed that the variables
like, glucose, peptone, yeast extract, malt extract and NaCl have significant effect on hydrogen yield and are selected for further optimization to attain a maximum yield (Fig. 2).
Fig. 1 - Parity plot between the experimental and predicted values of hydrogen yield
(α-ProteobacteriumAUChE 103) from Plackett–Burman design
y = 0.9955x
0.02
0.12
0.22
0.32
0.42
0.52
0.62
0.72
0.02 0.12 0.22 0.32 0.42 0.52 0.62 0.72
H2-Exp (mol H2/mol glucose)
H2-P
red
(m
ol
H2/m
ol
glu
co
se)
Hydrogen-Pred (mol H2/mol
glucose)
Linear (Hydrogen-Pred (mol
H2/mol glucose))
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 197
Table 3.1 Variables showing medium components used in Plackett-Burman design for the production of
hydrogen usingα-ProteobacteriumAUChE 103
Variables Low level High level
(–) values
(g/L) (+) values
(g/L) Glucose 20 40
Peptone 1 5
Beef Extract 0.5 2.5
Malt extract 1 3
Yeast extract 2 10
KCl 1 9
NaCl 1 5
NH4Cl 1 9
ZnCl2 0.5 5
KH2PO4 0.05 0.5
K2HPO4 .5 2.5
MnSo4.7H20 0.5 5
MgSo4.7H20 0.1 1
ZnSo4.7H20 0.1 5
FeSo4.7H20 0.1 1
ZnSo4.7H20
NH4Cl
KCl
KH2PO4
MgSo4.7H20
ZnCl2
Beefextract
MnSo4.7H20
K2HPO4
FeSo4.7H20
Yeastextract
NaCl
Peptone
Maltextract
Glucose
876543210
Term
Standardized Effect
2.776
Fig. 2 - Pareto’s Chart for (α-ProteobacteriumAUChE 103) screening of media constituents by Plackett-
Burman design
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 198
Table 3.2Twenty run Plackett-Burman design matrix (α-ProteobacteriumAUChE 103) for fifteen variables with the experimental and predicted
H2 yields
Glucose(A); Peptone(B); Beef extract(C); Malt extract(D); Yeast extract(E); KCl(F); NaCl(G); NH4Cl(H); ZnCl2(J); KH2PO4(K); K2HPO4(L);
MnSO4.7H2O(M); MgSO4.7H2O(N); ZnSO4.7H2O(O); and FeSO4.7H2O(P)
Runs A B C D E F G H J K L M N O P
Hydrogen yield
(mol H2/mol
glucose)
Exp. Pred.
1 -1 -1 1 -1 1 -1 1 1 1 1 -1 -1 1 1 -1 0.670 0.7229
2 1 -1 1 1 1 1 -1 -1 1 1 -1 1 1 -1 -1 0.023 0.0170
3 1 -1 1 -1 1 1 1 1 -1 -1 1 1 -1 1 1 0.262 0.2919
4 -1 1 1 -1 -1 -1 -1 1 -1 1 -1 1 1 1 1 0.719 0.6671
5 1 -1 -1 -1 -1 1 -1 1 -1 1 1 1 1 -1 -1 0.232 0.2720
6 -1 1 1 -1 1 1 -1 -1 -1 -1 1 -1 1 -1 1 0.673 0.6560
7 -1 1 -1 1 1 1 1 -1 -1 1 1 -1 1 1 -1 0.638 0.6550
8 -1 -1 -1 1 -1 1 -1 1 1 1 1 -1 -1 1 1 0.432 0.4069
9 1 1 1 1 -1 -1 1 1 -1 1 1 -1 -1 -1 -1 0.673 0.6411
10 1 -1 -1 1 1 -1 1 1 -1 -1 -1 -1 1 -1 1 0.021 0.0150
11 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 0.564 0.5939
12 -1 -1 -1 -1 1 -1 1 -1 1 1 1 1 -1 -1 1 0.696 0.6512
13 1 -1 1 1 -1 -1 -1 -1 1 -1 1 -1 1 1 1 0.139 0.1641
14 1 1 -1 -1 -1 -1 1 -1 1 -1 1 1 1 1 -1 0.720 0.6949
15 -1 1 1 1 1 -1 -1 1 1 -1 1 1 -1 -1 -1 0.531 0.5629
16 1 1 -1 -1 1 1 -1 1 1 -1 -1 -1 -1 1 -1 0.464 0.4192
17 1 1 -1 1 1 -1 -1 -1 -1 1 -1 1 -1 1 1 0.022 0.0759
18 -1 -1 1 1 -1 1 1 -1 -1 -1 -1 1 -1 1 -1 0.494 0.4641
19 1 1 1 -1 -1 1 1 -1 1 1 -1 -1 -1 -1 1 0.696 0.7279
20 -1 1 -1 1 -1 1 1 1 1 -1 -1 1 1 -1 1 0.464 0.5020
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 199
Table 3.3 Estimated effects and coefficients of the Plackett–Burman design for α-
ProteobacteriumAUChE 103
Terms Effect Coeffi. SE
Coeffi. T P
Constant
0.4569 0.01827 25.01 0.000
A. -0.2627 -0.1313 0.01827 -7.19 0.002
B. 0.2067 0.1033 0.01827 5.66 0.005
C. 0.0625 0.0312 0.01827 1.71 0.162
D. -0.2257 -0.1129 0.01827 -6.18 0.003
E. -0.1131 -0.0566 0.01827 -3.10 0.036
F. -0.0381 -0.0190 0.01827 -1.04 0.356
G. 0.1535 0.0768 0.01827 4.20 0.014
H. -0.0195 -0.0097 0.01827 -0.53 0.622
J. 0.0533 0.0267 0.01827 1.46 0.218
K. 0.0469 0.0234 0.01827 1.28 0.269
L. 0.0855 0.0428 0.01827 2.34 0.079
M. -0.807 -0.0403 0.01827 -2.21 0.092
N. -0.0533 -0.0267 0.01827 -1.46 0.218
O. -0.0013 -0.0007 0.01827 -0.04 0.973
P. -0.0881 -0.0440 0.01827 -2.41 0.073
Following screening, response surface methodology using CCD was applied to determine the optimal
levels of the significant variables. The coded and the actual values of the five significant medium components used in the design are presented in Table 3.4.The data obtained from the five level central composite design
matrix were used to develop models in which the dependent variable Y (hydrogen yield, mol H2/mol glucose)
was obtained as the sum of the contributions of the independent variables through second order polynomial
equation and interaction terms. The hydrogen yield obtained from experiments and from the model predictions are given in Table 3.5. The correlation coefficient, R
2, between the experimental and predicted data was 0.987,
revealing that 98.7% of experimental data of the hydrogen production was compatible with the data predicted
by the model. The parity plot shows a satisfactory correlation between the experimental and predicted values of the hydrogen production (Fig.3). The experimental results suggest that the minimum and maximum values of
hydrogen yield obtained were0.551mol H2/mol glucose and 0.791mol H2/mol glucose for Run No.50 and Run
No.34 respectively.
Regression analysis of the second order polynomial model for hydrogen production isgiven in Table
3.6. ANOVA indicated that, the linear and quadratic effects of glucose, peptone, malt extract and NaCl, and
interaction effects of glucose-peptone, glucose-malt extract, peptone-sodium chloride and yeast extract-sodium chloride concentration were significant. The production of hydrogen could be predicted by the model:
Y= 0.789 - 0.0038X1 + 0.001X2 + 0.015X3 + 0.0038X4 - 0.009X5 – 0.027X12 – 0.018X2
2
– 0.033X32 – 0.019X4
2 – 0.021X5
2 – 0.018X1X2 + 0.006 X1X3 + 0.012 X1X4 + 0.005X1X5 + 0.014 X2X3 + 0.014
X2X4 – 0.005 X2X5 –0.008X3X4 – 0.006 X3X5 + 0.0068X4X5………. (3.1)
where X1, the glucose concentration (g/L); X2, peptone (g/L); X3, malt extract concentration (g/L); X4, yeast
extract concentration (g/L) and X5, the sodium chloride concentration (g/L).
Hydrogen yield for different levels of variables was predicted from the respective contour plots (Fig. 4
(a-j)). Each contour plot represents an infinite number of combinations of the two test variables with the other
three maintained at their respective zero levels. There was a relative significant interaction between every two variables and the maximum predicted yield was indicated by the surface confined in the smallest ellipse in the
contour plot.
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 200
The optimum values obtained by solving the second degree polynomial equation for α-
ProteobacteriumAUChE 103 are glucose, 19.25g/L; peptone, 5.64g/L; malt extract, 1.64g/L; yeast extract, 3.16g/L and NaCl,4.312g/L.
Table 3.4Codes and actual levels of the independent variables for design of experiment for α-
ProteobacteriumAUChE 103 H2 fermentation
Table 3.5Five level factorial central composite design with experimental and predicted values of
hydrogen yield (mol H2/mol glucose) for α-ProteobacteriumAUChE 103 H2 fermentation
Ru
n
No.
Coded values Hydrogen yield (mol H2/mol glucose)
X1 X2 X3 X4 X5 Exp. Pred.
1 1.00 1.00 -1.00 -1.00 1.00 0.571 0.566
2 1.00 1.00 1.00 -1.00 1.00 0.643 0.637
3 0.00 0.00 0.00 2.38 0.00 0.696 0.686
4 0.00 0.00 0.00 0.00 0.00 0.790 0.789
5 1.00 1.00 -1.00 -1.00 -1.00 0.584 0.581
6 1.00 -1.00 1.00 1.00 1.00 0.668 0.683
7 -1.00 1.00 1.00 -1.00 1.00 0.678 0.683
8 0.00 0.00 0.00 0.00 0.00 0.790 0.789
9 -1.00 -1.00 1.00 1.00 -1.00 0.629 0.622
10 -1.00 1.00 1.00 1.00 1.00 0.702 0.694
11 -1.00 -1.00 -1.00 -1.00 -1.00 0.673 0.681
12 0.00 0.00 0.00 0.00 0.00 0.789 0.789
13 -1.00 -1.00 1.00 -1.00 -1.00 0.705 0.699
14 -1.00 1.00 1.00 1.00 -1.00 0.728 0.732
15 0.00 2.38 0.00 0.00 0.00 0.694 0.685
16 1.00 -1.00 1.00 -1.00 -1.00 0.701 0.707
17 0.00 0.00 0.00 -2.38 0.00 0.672 0.668
18 -1.00 -1.00 1.00 1.00 1.00 0.598 0.605
19 1.00 -1.00 -1.00 -1.00 1.00 0.662 0.665
20 1.00 -1.00 -1.00 1.00 -1.00 0.672 0.669
21 0.00 0.00 0.00 0.00 -2.38 0.695 0.687
22 -1.00 1.00 1.00 -1.00 -1.00 0.740 0.749
23 2.38 0.00 0.00 0.00 0.00 0.633 0.623
24 1.00 -1.00 -1.00 -1.00 -1.00 0.652 0.664
25 -1.00 -1.00 1.00 -1.00 1.00 0.643 0.654
26 1.00 -1.00 1.00 1.00 -1.00 0.680 0.679
27 1.00 -1.00 1.00 -1.00 1.00 0.691 0.684
Independent
variables Symbols
Coded levels
-2 -1 0 +1 +2
Glucose, (g/L) X1 10 15 20 25 30 Peptone, (g/L) X2 1 3 5 7 9 Malt extract, (g/L) X3 0.5 1.0 1.5 2.0 2.5 Yeast extract, (g/L) X4 1 2 3 4 5 NaCl, (g/L) X5 1 3 5 7 9
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 201
Ru
n
No.
Coded values Hydrogen yield (mol H2/mol glucose)
X1 X2 X3 X4 X5 Exp. Pred.
28 -2.38 0.00 0.00 0.00 0.00 0.644 0.640
29 1.00 1.00 1.00 -1.00 -1.00 0.678 0.682
30 0.00 0.00 0.00 0.00 0.00 0.790 0.789
31 1.00 -1.00 -1.00 1.00 1.00 0.701 0.698
32 -1.00 1.00 -1.00 1.00 -1.00 0.680 0.690
33 -1.00 -1.00 -1.00 1.00 1.00 0.643 0.645
34 0.00 0.00 0.00 0.00 0.00 0.791 0.789
35 1.00 1.00 1.00 1.00 -1.00 0.701 0.714
36 1.00 1.00 -1.00 1.00 1.00 0.645 0.653
37 0.00 0.00 0.00 0.00 0.00 0.790 0.789
38 1.00 1.00 -1.00 1.00 -1.00 0.648 0.646
39 0.00 0.00 0.00 0.00 0.00 0.788 0.789
40 -1.00 -1.00 -1.00 -1.00 1.00 0.674 0.661
41 0.00 0.00 2.38 0.00 0.00 0.643 0.636
42 -1.00 -1.00 -1.00 1.00 -1.00 0.626 0.637
43 -1.00 1.00 -1.00 -1.00 1.00 0.623 0.632
44 0.00 0.00 0.00 0.00 0.00 0.789 0.789
45 0.00 0.00 0.00 0.00 0.00 0.789 0.789
46 1.00 1.00 1.00 1.00 1.00 0.695 0.696
47 0.00 -2.38 0.00 0.00 0.00 0.682 0.678
48 -1.00 1.00 -1.00 -1.00 -1.00 0.685 0.674
49 0.00 0.00 0.00 0.00 2.38 0.648 0.643
50 0.00 0.00 -2.38 0.00 0.00 0.551 0.554
51 -1.00 1.00 -1.00 1.00 1.00 0.676 0.676
52 0.00 0.00 0.00 0.00 0.00 0.790 0.789
Table 3.6Results of the regression analysis of second order polynomial model for theoptimization of
hydrogen production using α-ProteobacteriumAUChE 103
Term
constant Regression
coefficient Std. deviation
T-statistics P-value
Intercept 0.788605 0.002844 277.299 <0.001 X1 -0.003651 0.001375 -2.655 0.012 X2 0.001536 0.001375 1.117 0.272 X3 0.015171 0.001375 11.035 <0.001 X4 0.003857 0.001375 2.805 0.009 X5 -0.009275 0.001375 -6.746 <0.001 X1X1 -0.027689 0.001182 -23.433 <0.001 X2X2 -0.018950 0.001182 -16.038 <0.001 X3X3 -0.033250 0.001182 -28.140 <0.001 X4X4 -0.019656 0.001182 -16.635 <0.001 X5X5 -0.021863 0.001182 -18.503 <0.001 X1X2 -0.018875 0.001600 -11.799 <0.001 X1X3 0.006250 0.001600 3.907 0.001 X1X4 0.012125 0.001600 7.579 <0.001 X1X5 0.005250 0.001600 3.282 0.003 X2X3 0.014438 0.001600 9.025 <0.001 X2X4 0.014938 0.001600 9.337 <0.001
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 202
X2X5 -0.005437 0.001600 -3.399 0.002 X3X4 -0.008313 0.001600 -5.196 <0.001 X3X5 -0.006188 0.001600 -3.868 <0.001 X4X5 0.006812 0.001600 4.258 <0.001
Table 3.7Analysis of the variance (ANOVA) for the fitted quadratic polynomial model for the production
of hydrogen using α-ProteobacteriumAUChE 103
Sources of
variation Sum of
squares Degrees of
freedom (DF)
Mean
square
(MS) F- value P-value
Regression 0.191106 20 0.191106 116.68 <0.001 Linear 0.015025 5 0.015025 36.69 <0.001 Square 0.138167 5 0.138167 337.42 <0.001 Interaction 0.037915 10 0.037915 46.30 <0.001 Residual
Error 0.002539 31 0.002539 -
Lack-of-Fit 0.002534 22 0.002534 230.39 <0.001 Pure Error 0.000004 9 0.000004 - - Total 0.193645 51 - - -
Table 3.8Optimum values of the variables obtained from regression equation
Independent variables Optimum value (coded) Optimum value (real) (g/L)
Glucose -0.1683 19.25 Yeast extract 0.0240 3.160 Malt extract 0.3125 1.640 Peptone 0.3125 5.640 NaCl -0.3125 4.312
Fig.3 - Parity plot showing the distribution of experimental versus predicted values of hydrogen yield (α-
ProteobacteriumAUChE 103) by RSM
y = 0.9988x
0.56
0.59
0.62
0.65
0.68
0.71
0.74
0.77
0.56 0.59 0.62 0.65 0.68 0.71 0.74 0.77
H2-Exp (mol H2/mol glucose)
H2-P
red
(m
ol
H2/m
ol
glu
co
se)
Hydrogen-Pred (mol H2/mol
glucose)
Linear (Hydrogen-Pred (mol
H2/mol glucose))
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 203
(a) (b)
(c) (d)
(e) (f)
-2.37
-1.19
0.00
1.19
2.37
-2.37
-1.19
0.00
1.19
2.37
0.2
0.2475
0.295
0.3425
0.39
H
yd
rog
en
Yie
ld
C: Beef extract E: NaCl -2.37
-1.19
0.00
1.19
2.37
-2.37
-1.19
0.00
1.19
2.37
0.23
0.2675
0.305
0.3425
0.38
H
yd
rog
en
Yie
ld
D: Yeast extract E: NaCl
-2.31
-1.16
0.00
1.15
2.31
-2.31
-1.16
0.00
1.15
2.31
0.43
0.52
0.61
0.7
0.79
H
yd
ro
ge
n Y
ield
A: Glucose B: Peptone
-2.31
-1.16
0.00
1.15
2.31
-2.31
-1.16
0.00
1.15
2.31
0.38
0.485
0.59
0.695
0.8
H
yd
ro
ge
n Y
ield
A: Glucose C: Malt extract
-2.31
-1.16
0.00
1.15
2.31
-2.31
-1.16
0.00
1.15
2.31
0.45
0.535
0.62
0.705
0.79
H
yd
ro
ge
n Y
ield
A: Glucose D: Yeast extract -2.31
-1.16
0.00
1.15
2.31
-2.31
-1.16
0.00
1.15
2.31
0.4
0.5
0.6
0.7
0.8
H
yd
ro
ge
n Y
ield
B: Peptone C: Malt extract
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 204
(g) (h)
(i) (j)
Fig. 4 (a-j) - 3D plots showing the interactive effects between significant variables on hydrogen yield in α-
ProteobacteriumAUChE 103 batch fermentation
3.4 Process parameter optimization using Box-Behnken design
The Box-Behnken design was used to optimize the important process parameters namely, initial pH of the
medium, temperature and time of fermentation.The coded and the actual values of the parameters used in the
design are presented in Table 3.9. The design matrix which consists of 15 experimental runs was constructed, in order to arrive at a second order polynomial equation to predict the hydrogen fermentation system. The design
matrix and their corresponding experimental and the predicted values are given in Table 3.10. The highest
hydrogen yield of 0.89 mol H2/mol glucose was obtained for the experimental runs of 1, 4 and 9. The results
were analyzed using the analysis of variance (ANOVA) and the estimated coefficients are presented in Table 3.11. The hydrogen yield using α-ProteobacteriumAUChE 103can be expressed in terms of the following
regression equation;
Y=0.887 + 0.049A + 0.0137B + 0.007C + 3.25AB + 0.015AC + 0.01BC - 0.07A2 -
0.09B2 - 0.07C
2 ……… (3.2)
where, A, pH; B, time and C, the temperature.
The multiple correlation coefficient, R2, obtained from the ANOVA was 0.9998, which indicate that the
model is capable of explaining 99.98% of the variation in response. This is supported by the parity plot between
the experimental and predicted hydrogen yield (Fig. 5).Three dimensional surface plots are drawn to determine
the optimum values and the interactive effect of the three process parameters (Fig. 6 (a-c)). The experimental
-2.31
-1.16
0.00
1.15
2.31
-2.31
-1.16
0.00
1.15
2.31
0.48
0.5575
0.635
0.7125
0.79
H
yd
rog
en
Yie
ld
A: Glucose E: NaCl -2.31
-1.16
0.00
1.15
2.31
-2.31
-1.16
0.00
1.15
2.31
0.52
0.5875
0.655
0.7225
0.79
H
yd
rog
en
Yie
ld
B: Peptone E: NaCl
-2.31
-1.16
0.00
1.15
2.31
-2.31
-1.16
0.00
1.15
2.31
0.44
0.53
0.62
0.71
0.8
H
yd
ro
ge
n Y
ield
C: Malt extract E: NaCl -2.31
-1.16
0.00
1.15
2.31
-2.31
-1.16
0.00
1.15
2.31
0.5
0.5725
0.645
0.7175
0.79
H
yd
ro
ge
n Y
ield
D: Yeast extract E: NaCl
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 205
results suggest that the maximum values of hydrogen yield (0.89 mol H2/mol glucose) were obtained for the
runs with the central points.The optimum values obatained by solving the second degree polynomial equation are as follows: pH, 7.0; temperature, 34.5
oC and fermentation time, 42.5h.
Fig. 5 - Parity plot between the experimental and predicted values of hydrogen yield (α-
ProteobacteriumAUChE 103) by Box-Behnken design
Table 3.9The coded and actual values of the variables used in the design
Parameters Coded values
-1 0 1
pH (A) 6.5 7.0 7.5
Time (h) (B) 40 42 44
Temperature (oC) (C) 32 34 36
Table 3.10 Three level Box-Behnken design matrix for the optimization of significant process parameters
in α-Proteobacterium AUChE 103 H2 fermentation
Run No.
pH Temperature Time
Hydrogen yield (mol H2/mol glucose)
Exp. Pred.
1 0.00
0.00
0.00 0.889 0.888
2 -1.00 0.00 -1.00 0.700 0.701
3 -1.00 0.00 1.00 0.685 0.683
4 0.00 0.00 0.00 0.886 0.888
5 1.00 0.00 1.00 0.812 0.812
6 0.00 1.00 1.00 0.747 0.748
7 1.00 1.00 0.00 0.785 0.785
8 0.00 1.00 -1.00 0.715 0.713
y = 1.0002x
0.66
0.71
0.76
0.81
0.86
0.66 0.71 0.76 0.81 0.86
H2-Exp (mol H2/mol glucose)
H2-P
red
(m
ol
H2/m
ol
glu
co
se)
Hydrogen-Pred (mol H2/mol glucose)
Linear (Hydrogen-Pred (mol H2/mol
glucose))
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 206
9 0.00 0.00 0.00 0.888 0.888
10 -1.00 1.00 0.00 0.680 0.681
11 -1.00 -1.00 0.00 0.660 0.660
12 1.00 0.00 -1.00 0.764 0.766
13 1.00 -1.00 0.00 0.752 0.751
14 0.00 -1.00 1.00 0.698 0.700
15 0.00 -1.00 -1.00 0.707 0.706
Table 3.11 Results of the ANOVA of the process parameter optimization data in α-
ProteobacteriumAUChE 103 H2 fermentation using Box-Behnken design of experiments
Source Sum of Squares
Degrees of
Freedom (DF) Mean Square
F Value
P-value Prob> F
Model 0.12 9 0.013 3331.77 < 0.0001 pH (A) 0.019 1 0.019 4878.74 < 0.0001 Temperature (B)
1.513×10-3 1 1.513×10
-3 392.13 < 0.0001
Time (C) 3.920×10-4 1 3.920×10
-4 101.63 < 0.0001 AB 4.225×10
-5 1 4.225×10-5 10.95 0.0129
AC 9.923×10-4 1 9.923×10
-4 257.25 < 0.0001 B C
4.202×10-4 1 4.202×10
-4 108.95 < 0.0001
A^2 0.022 1 0.022 5678.61 < 0.0001 B^2 0.039 1 0.039 9981.91 < 0.0001 C^2 0.023 1 0.023 6079.10 < 0.0001 Residual 2.700×10
-5 7 3.857×10-6
Lack of fit 1.900×10-5 3 6.333×10
-6 3.17 0.1473 Pure error 8.000×10
-6 4 2.000×10-6
Total 0.12 16
Table 3.12 Estimated coefficient values for α-ProteobacteriumAUChE 103 H2 fermentation
Factor Coefficient Estimate
Degrees of
freedom
(DF)
Standard error
95% CI Low
95% CI High
VIF
Intercept 0.89 1 8.783×10-4 0.88 0.89
pH (A) 0.049 1 6.944×10-4 0.047 0.050 1.00
Temperature
(B) 0.014 1 6.944×10
-4 0.012 0.015 1.00
Time (C) 7.000×10-3 1 6.944×10
-4 5.358×10-3 8.642×10
-3 1.00
AB 3.250×10-3 1 9.820×10
-4 9.280×10-4 5.572×10
-3 1.00
AC 0.016 1 9.820×10-4 0.013 0.018 1.00
BC 0.010 1 9.820×10-4 7.928×10
-3 0.013 1.00
A^2 -0.072 1 9.571×10-4 -0.074 -0.070 1.01
B^2 -0.096 1 9.571×10-4 -0.098 -0.093 1.01
C^2 -0.075 1 9.571×10-4 -0.077 -0.072 1.01
(a) (b)
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 207
(c)
Fig. 6 (a-c)- 3D plots showing the interactive effects between the significant process parameters on
hydrogen yield in α-ProteobacteriumAUChE 103 batch fermentation
3.5 Effect of different substrates
Different substrates such as, glucose, sucrose, sugarcane bagasse hydrolysate, sorghum stalk
hydrolysate and cellulose were tested for α-ProteobacteriumAUChE 103 for its ability to utilize variety of
sugars and its hydrogen synthesizing ability. The cheaper lignocellulosic substrate, sorghum stalk hydrolysate
has produced a better hydrogen yield (0.91 mol H2/mol glucose) compared to all synthetic carbon and the other lignocellulosic substrate (sugarcane bagasse hydrolysate) (Fig. 7). In order to understand the kinetics of α-
ProteobacteriumAUChE 103 hydrogen fermentation under different initial substrate concentration, the
experiments were performed at different initial hydrolysate concentrations ranging between 0.5% and 3% (w/v) (glucose equivalent). High initial substrate concentration may play an important role in hydrogen production
rates and yields [12-15]. However, fermenting microorganisms can also have limited tolerance to increased
substrate loading [16-17]. The rate of total gas collection was almost constant throughout the fermentation periods. The percentage hydrogen content in the gas mixture almost doubles during the period of 30 hours to 36
hours. The rate of hydrogen production increases linearly up to 30 hours of fermentation, after that, almost it
doubles up (Fig. 8). The maximum volumetric hydrogen productivity of 8.10 ml/Lh was obtained at a
fermentation period of 42 hours with 1.5% (glucose equivalent) initial substrate concentration.
-1.00
-0.50
0.00
0.50
1.00
-1.00
-0.50
0.00
0.50
1.00
0.66
0.72
0.78
0.84
0.9
H
yd
ro
ge
n Y
ield
A: pH B: Time -1.00
-0.50
0.00
0.50
1.00
-1.00
-0.50
0.00
0.50
1.00
0.68
0.735
0.79
0.845
0.9
H
yd
ro
ge
n Y
ield
A: pH C: Temperature
-1.00
-0.50
0.00
0.50
1.00
-1.00
-0.50
0.00
0.50
1.00
0.69
0.74
0.79
0.84
0.89
H
yd
ro
ge
n Y
ield
B: Time C: Temperature
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 208
Fig. 7 - Effect of different substrates on hydrogen yield in α-ProteobacteriumAUChE 103 batch cultures
Fig. 8 - Time profile of hydrogen yield for different substrate concentrations inα-ProteobacteriumAUChE
103 batch culture
4. Conclusions
The study demonstrated the optimization of media components and process parameters for the production of hydrogen using the isolated strain, α-ProteobacteriumAUChE 103.An optimum hydrolysis time
and acid concentration is required for the proper hydrolysis of sorghum stalk, above and below which the
glucose yield was low. A suitable sorghum stalk hydrolysate in the fermentation medium is essential to get higher hydrogen yields; however, an excessively high or low concentration of sorghum stalk hydrolysate will
affects the growth of the organism which will reduce the yield.The optimum concentration of medium
components for α-ProteobacteriumAUChE 103 hydrogen production was found to be: glucose, 19.25g/L; yeast extract, 3.046g/L; malt extract, 1.64g/L; peptone, 5.640 and NaCl, 4.312g/L. Box-Behnken design was used to
test the importance of process parameters on hydrogen production. The optimized values of process parameters
for hydrogen production were as follows: pH, 7.0; temperature, 34.5oC and fermentation time, 42.5h. A highest
hydrogen yield of 0.91 mol H2/mol glucose was achieved, when sorghum stalk hydrolysate (equivalent to 1.5% (w/v)) was used as the carbon source.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0.5 1 1.5 2 3
Concentrations
Hy
dro
gen
yie
ld
(mo
l H
2/m
ol
glu
co
se)
Glucose % (w/v) Sucrose % (w/v)
Cellulose % (w/v) Bagasse hydrolysate % (v/v)Sorghum stalk hydrolysate % (v/v)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 6 12 18 24 30 36 42 48
Time (h)
Hy
dro
gen
yie
ld
(mo
l H
2/m
ol
glu
cose
)
0.5% (v/v)
1 %(v/v)
1.5% (v/v)
2 %(v/v)
3 %(v/v)
T.R.Manikkandan /International Journal of ChemTech Research, 2019,12(4): 194-209. 209
References
1. Hsia S, Chou Y. Optimization of Biohydrogen Production with Biomechatronics. J Nanomaterials
2014; 1-11. 2. Nath K, das D. Improvement of fermentative hydrogen production: Various approaches. Appl.
Microbial Biotechnol 2004;65:520-9.
3. Kapdan I K, Kargi F. Bio-hydrogen production from waste material. Enzyme and Microbial Biotechnol
2006;38:569-82. 4. Winter C J. In to the hydrogen economy milestones. Int J Hydrogen Energy 2005;30:681-5.
5. Nath K, Das D. Hydrogen from biomass. Current Science 2003;85.
6. Rollin JA, Martin del Campo J, Myung S, Sun F, You C, Bakovic A, Castro R, Chandrayan SK, Wu CH, Adams MWW, Senger RS, Zhang YHP. High-yield hydrogen production from biomass by in vitro
metabolic engineering: Mixed sugars coutilization and kinetic modeling.Proc. Natl. Acad. Sci2015; 112
(16): 4964–9 7. Saxena R C, Adhikari D K, Goyal H B. Biomass based energy fuel through biochemical routes: A
review. Renewable and Sustainable Energy Reviews 2018;13:167-78.
8. Patrick C H. Fermentative hydrogen production: Principles, progress and prognosis. Int J Hydrogen
Energy 2009; 34: 7379-89. 9. Xing Y, Ma H, Fan Y, Hou H, Chen J.Cellulose-hydrogen production from corn stalk biomass by
anaerobic fermentation.Chinese Sci. Bull.,2009; 54 (8): 1434–41.
10. Nissila ME, Lay C, Puhakka JA. Dark fermentative hydrogen production from lignocellulosic hydrolyzates - A review.Biomass and Bioenergy2014; 67: 145–59.
11. Rai PK, Singh SP, Asthana RK, Singh S. Biohydrogen production from sugarcane bagasse by
integrating dark- and photo-fermentation.Bioresour. Technol2017; 152: 140–6.
12. Kumar N, Das D. Continuous hydrogen production by immobilized Enterobacter cloacae IIT-BT 08 using lignocellulosic materials as solid matrices. Enzyme and Microbial Technol 2001;29:280-7.
13. Lacis LS, Lawford HG. Ethanol production from xylose by Thermoanaerobacterethanolicus in batch
and continuous culture. Archives of Microbiology 1998;150:48-55. 14. Van Ginkel S, Sung S. Biohydrogen production as a function of pH and substrate concentration.
Environ SciTechnol 2001;35:4726-30.
15. Sommer P, Georgieva T, Ahring BK. Potential for using thermophilic anaerobic bacteria for bioethanol production from hemicelluloses. Biochemical Society Transactions 2004;32:283-9.
16. Van Ginkel S, Logan BE. Inhibition of bio-hydrogen production by undissociated acetic and butyric
acids. Environ SciTechnol 2005;39:9351-6.
17. Olsson L, Hahn-Hagerdahl B. Fermentation of lignocellulosichydrolysates for ethanol production. EnzMicrobTechnol 1996;18:312-31.
*****